agnosticmantis

9 ms·

91.
by agnosticmantis 2y ago
It's amusing how Sutskever kept musking Musk over the years (overpromising with crazy deadlines and underdelivering): In 2017 he wrote "Within the next three years, robotics should be completely solved, AI should solve a long-stan
92.
by agnosticmantis 2y ago
From Musk's email: "Frankly, what surprises me is that the AI community is taking this long to figure out concepts. It doesn’t sound super hard. High-level linking of a large number of deep nets sounds like the right approach or a
93.
by agnosticmantis 2y ago
Related to the Hyperreal numbers mentioned in the article is the class of Surreal numbers which have many fun properties. There's a nice book describing them authored by Don Knuth.
94.
5-Day Gen AI Intensive Course with Google Learn Guide(kaggle.com)
2 points by agnosticmantis 2y ago|0 comments
95.
by agnosticmantis 2y ago
I found that I learned a lot about RCTs by going beyond RCTs and reading about causal inference. You learn why each assumption is important when it's broken. 'Causal Inference: What If' is a nice intro and freely available:
96.
by agnosticmantis 2y ago
Now OpenAI only needs to figure out how to do Search profitably instead of burning VC money. Only then will the comparison with profitable Search engines be sensible. Before enshitification all VC-backed products are a delight to use, but a
97.
by agnosticmantis 2y ago
These papers don’t explain how pertained LLMs learn in-context, because the simplified models in these papers are either pretrained for the same task that’s tested in-context, or the weights are handpicked by humans to do GD at inference ti
98.
Oakland's Psychedelic Mushroom Church(oaklandside.org)
25 points by agnosticmantis 2y ago|4 comments
99.
AI and Stress(docs.google.com)
1 points by agnosticmantis 2y ago|0 comments
100.
by agnosticmantis 2y ago
I’m guessing it was the second attempt that really stuck.
101.
by agnosticmantis 2y ago
How GPU-friendly is this class of models?
102.
by agnosticmantis 2y ago
For a moment I thought Andrej Karpathy was the voice actor for the AI Assistant in the game.
103.
by agnosticmantis 2y ago
If I didn't know this was from 2007 I'd have said it shows that he's been hanging out with OpenAI kids. (He worked there working on watermarking and stuff.) (Altman having said they're working on an AI OS or something.)
104.
The official PyTorch implementation of Google's Gemma models(github.com)
2 points by agnosticmantis 3y ago|0 comments
105.
by agnosticmantis 3y ago
> However, the idea is that often a lot of the probability mass - an amount that is not small - will be concentrated around the maximum likelihood estimate, and so that's why it makes a good estimate, and worth using. This is a Baye
106.
ChatGPT but It's in 2006(youtube.com)
2 points by agnosticmantis 3y ago|0 comments
107.
V-JEPA: Video Joint Embedding Predictive Architecture (V-JEPA) Model(ai.meta.com)
68 points by agnosticmantis 3y ago|7 comments
108.
by agnosticmantis 3y ago
According to [0]: - Books3 dataset - 700B text-image pairs from Laion-2B-en, filtered to only keep images with at least 256 resolution - 400M text-image pairs from COYO-700M, filtered to only keep images with at least 256 resolution - 10M t
109.
by agnosticmantis 3y ago
Dating apps have very peculiar dynamics (e.g. you need to somehow get women on the app and men will follow automatically). Also women may be conservative and they might not want to join the latest dating app that ranks 50th on the App Store
110.
by agnosticmantis 3y ago
What I really like and admire about Euler is how masterfully he handled infinities and infinitesimals to arrive at correct conclusions (the vast majority of times anyway) even though analysis hadn’t been made rigorous yet (by Cauchy and Wei
111.
by agnosticmantis 3y ago
Did you pay for the video proportional to the value that you received? Most people don’t, so why expect the services to not get worse, and be forever subsidized by another cash cow?
112.
by agnosticmantis 3y ago
What are the severance packages like this time?
113.
by agnosticmantis 3y ago
Reminds me of John Snow's discovery and demonstration of the cause of cholera, which I learned about in the context of casual inference in this excellent paper by statistician David Freedman: https://psychology.okstate.edu&#
114.
by agnosticmantis 3y ago
My mind was blown recently discovering Tubi with so many great movies (many of them classics and highly rated) for free (with occasional ads, of course) and I regret having paid for Netflix and other services in the past.
115.
NotebookLM: Personalized AI, grounded in the information you trust(notebooklm.google)
2 points by agnosticmantis 3y ago|0 comments
116.
by agnosticmantis 3y ago
https://notebooklm.google
117.
Labs.Google(labs.google)
130 points by agnosticmantis 3y ago|105 comments
118.
by agnosticmantis 3y ago
Schrödinger’s Charlatan: he’s both a genius and the most naïve when it comes to FSD, at the same time. No sir, he’s been making fraudulent claims about present capabilities not just future: convoy tech 10 times safer than humans, NOW.
119.
23andMe trying to trick customers into accepting TOS change, to prevent lawsuit(hachyderm.io)
43 points by agnosticmantis 3y ago|4 comments
120.
by agnosticmantis 3y ago
Can we (folks who didn't use 23&me but may still be affected because our relatives might have) file a class action lawsuit? We haven't signed any licensing agreements with 23&me waiving our privacy, so presumably we still

More ›